Inferring angiosperm phylogeny from EST data with widespread gene duplication

Table 2 Effects of hit fraction threshold on cluster assembly. Bold indicates the threshold chosen for the current study.

Hit fraction^a	Clusters^b	Singletons^c	Phylogenetically informative clusters^d	Max size^e	TCs in phylogenetically informative clusters^f
0.0	39924	26782	4423	6565	54051
0.1	47798	32824	4079	1947	42406
0.2	57229	41327	3324	1362	29403
0.3	64691	48864	2561	330	21504
0.4	71333	56383	1876	117	15457
0.5	77564	63890	1340	98	10721
0.6	83435	71539	897	95	7105
0.7	88864	79122	577	94	4536
0.8	94296	87186	324	92	2529
0.9	99843	95975	103	89	872
1.0	105144	104860	1	6	6

^a Minimum proportion of sequence similarity based on BLAST's pairwise comparisons. The hit fraction determines whether a sequence is linked to another (if a pair is linked, they will be placed in the same cluster) and thus affects the level of heterogeneity within clusters and the number of assembled clusters. Original number of sequences is 105,453 TCs.
^b Total number of assembled clusters.
^c Number of single-sequence clusters.
^d Phylogenetically informative clusters for this study are those that include at least three species and at least four sequences.
^e Number of tentative consensus sequences (TCs) in the largest phylogenetically informative cluster.
^f Total TCs in all phylogenetically informative clusters.

ISSN: 2730-7182